Impact of Polymer Matrix on the Electromagnetic ...for high performance EMI shielding.2 12 13 15 22 While most reports point out that the structure and prop-erties of CNTs strongly

Delivered by Publishing Technology to: National Tsing Hua UniversityIP: 140.114.93.136 On: Thu, 25 Jul 2013 06:33:15

Copyright: American Scientific Publishers

RESEARCH

ARTIC

LE

Copyright © 2013 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofNanoscience and Nanotechnology

Vol. 13, 1120–1124, 2013

Impact of Polymer Matrix on the ElectromagneticInterference Shielding Performance for Single-Walled

Carbon Nanotubes-Based Composites

Jiajie Liang, Yi Huang∗, Ning Li, Gang Bai, Zunfeng Liu,Feng Du, Feifei Li, Yanfeng Ma, and Yongsheng Chen∗

Key Laboratory of Functional Polymer Materials and Center for Nanoscale Science and Technology,Institute of Polymer Chemistry, College of Chemistry, Nankai University, Tianjin 300071, China;

Faculty of Chemistry and Material Engineering, Wenzhou University,Zhejiang Province, Wenzhou, 325027, China

Composites of acrylonitrile butadiene styrene (ABS), epoxy and soluble cross-linked polyurethane(SCPU) with various loadings of single-walled carbon nanotubes (SWCNTs) were prepared.Their electromagnetic interference (EMI) shielding effectiveness (SE) in the frequency range of8.2–12.4 GHz (X band) was studied. Well-dispersed SWCNT composites were created in thesethree representative polymer matrixes. The choice of polymer matrix greatly affects the conductivity,percolation threshold, and EMI shielding properties of the SWCNT/polymer composites. EnhancedEMI SE performances were observed for the composites with better dispersed SWCNTs. Moreover,the EMI SE performances strongly correlated with SWCNT loading in the polymer matrix. The bestSWCNT dispersion was achieved in the epoxy matrix: 20–30 dB EMI SE was obtained with 15 wt%SWCNTs.

Keywords: Electromagnetic Interference Shielding, Single-Walled Carbon Nanotubes,Composite.

1. INTRODUCTION

Materials with electromagnetic interference (EMI) shield-ing effectiveness (SE) are in high demand for bothcommercial and military applications. Conventionally, inor-ganic magnetic or metallic particles are used as EMImaterials. However, their high specific gravity and poorprocessibility have limited their practical applications.Thus, the urgent need remains for EMI materials thatare relatively lightweight, structurally sound and flexibleand efficient in wide band-range shielding or absorption.Recently, nano-structured materials have generated signif-icant attention for these applications due to their manyunique chemical and physical properties.1–4 With a low spe-cific mass and excellent thermal, electrical and mechan-ical properties, carbon nanotubes (CNTs), including bothsingle-walled (SWCNTs) and multi-walled (MWCNTs)carbon nanotubes, have been studied for potential engi-neering applications in electronics, electrostatic dissipation,

∗Authors to whom correspondence should be addressed.

multilayer printed circuits, and conductive coatings.5–11

Their excellent electrical properties, small diameters andhigh aspect ratios,12–14 have made them an excellent optionfor high performance EMI shielding.2�12�13�15–22

While most reports point out that the structure and prop-erties of CNTs strongly influence the EMI shielding per-formance of CNT/polymer composites,1�15–17�19�23�24 fewreports focus on the influence of polymer matrix.20

Recently, we reported the impact of SWCNT structure(i.e., diameter, aspect ratio and wall integrity) on the EMIperformance of epoxy composites.23�24 In this work, westudied the influence of different polymer matrices onthe EMI SE performance of SWCNT/polymer compos-ites in the microwave band. SWCNT/polymer compositeswith varying SWCNT loading and three sets of represen-tative matrix polymers (thermoplastic acrylonitrile buta-diene styrene (ABS), thermosetting epoxy and solublecross-linked polyurethane (SCPU)) were prepared in thiswork. The results indicated that the polymer matrix, whichstrongly affects the dispersion state of SWCNTs, has greatimpact on the conductivity, percolation threshold, and EMI

1120 J. Nanosci. Nanotechnol. 2013, Vol. 13, No. 2 1533-4880/2013/13/1120/005 doi:10.1166/jnn.2013.5997



RESEARCH

ARTIC

LE

Liang et al. Impact of Polymer Matrix on the EMI Shielding Performance for SWCNTs-Based Composites

SE properties of the composites. The epoxy compositesdisplayed the best SWCNT dispersion and resulted in20–30 dB of EMI SE with 15 wt% SWCNT loading.

2. EXPERIMENTAL DETAILS

2.1. Preparation of SWCNT/Polymer Composites

SWCNTs were prepared in our laboratory with a modifiedarcing method at large scale using a Ni/Y catalyst.25 TheSWCNT/SCPU composites, SWCNT/ABS composites andSWCNT/Epoxy composites were prepared according toour previously works.23�24�26

2.2. Instruments and Measurements

Scanning electron microscopy (SEM) was performedon a Hitachi S-3500N scanning electron microscope.Direct-current (DC) conductivity of the SWCNT/polymercomposites was determined using the standard four-pointcontact method on rectangular sample slabs to eliminatethe contact-resistance effect. The data were collected witha Keithley SCS 4200. The EMI SE and permittivity data ofthe SWCNT composites were measured using 22.86 mm×10.16 mm×2 mm slabs (to fit waveguide sample holder),by an HP vector network analyzer (HP E8363B) in the8.2–12.4 GHz (X band) range. All ultrasonication process-ing was performed with a sonicator (Gongyi Yuhua Instru-ment Co., LTD., Model: KQ400B, 400 W).

3. RESULTS AND DISCUSSION

The EMI SE properties of composites are highly related tothe filler’s intrinsic conductivity, dielectric constant, aspectratio12�13 and dispersion state. And it is well known thatthe conductivity of conductor-insulator composites fol-lows the critical phenomena around threshold. Figure 1

Fig. 1. log10 DC conductivity (�� versus mass fraction (p) ofSWCNT/polymer composites (Red circle: SWCNT/SCPU, Blue trian-gle: SWCNT/ABS, Black square: SWCNT/epoxy) measured at roomtemperature.

shows the dc conductivity (�� of our SWCNT/SCPU,SWCNT/ABS and SWCNT/epoxy composites as a func-tion of the SWCNT mass fraction (p�. All compositesexhibited a change of over 10 orders of magnitude at dif-ferent SWCNT loadings, indicating the formation of thepercolating network.To date, different thresholds have been found for the

conductivity of SWCNT/polymer composites. Since lowerfilling fractions imply lower cost and smaller perturbationof bulk physical properties, it is crucial to have a lowfilling threshold for practical applications. It is well knownthat the conductivity of a conductor-insulator compositefollows the critical phenomena around the percolationthreshold (Eq. (1)):27

� ∝ ��−�c�� (1)

where � is the composite conductivity, � is the SWCNTvolume fraction, �c is the percolation threshold and � isthe critical exponent. Because the densities of the poly-mer and SWCNT are similar, we assume that the massfraction, p, and the volume faction, �, of the SWCNTs inthe polymer are similar. log–log plots (� vs. (p−pc�/pc�were generated using Eq. (1) and least-square fit to thedata near the threshold. The fit shows that the thresh-old volume pc for each set of composites was stronglybounded by the regions between the highest insulat-ing and lowest conducting points. The conductivity ofSWCNT/polymer composites agreed very well with thepercolation behaviour predicted by Eq. (1).In previous works, both theoretical and experimental

results for the percolation threshold (pc� of compositesdepended primarily on the filler’s aspect ratio, processingmethods and matrix.13�23�28 Table I summarizes the per-colation thresholds and critical exponents obtained usingEq. (1) for the three composites. The percolation thres-hold of SWCNT/epoxy composites was found to be quitelow (0.062 wt%) and in good agreement with previousstudies of SWCNT/polymer systems.29�30 This indicatedthat our processing method distributed the SWCNTs wellin the epoxy matrix. Higher values for the percolationthreshold were obtained for the SWCNT/ABS (0.599 wt%)and SWCNT/SCPU composites (3.34 wt%). Thus, thebest dispersion of SWCNTs was achieved in the epoxymatrix. The values of the critical exponent � were alsoin good agreement with the theoretical results for a per-colating rod network system.28�31 The conductivity of theSWCNT/polymer composites all dramatically increased at

Table I. Percolation thresholds, critical exponents, and correlation fac-tors for the three SWCNT/polymer composites.

SWCNTs Percolation Critical Correlationused threshold (wt%), pc exponent B factor R

SWCNT/SCPU 3�39 4�7 0.97SWCNT/ABS 0�599 2�95 0.99SWCNT/epoxy 0�062 2�68 0.98

J. Nanosci. Nanotechnol. 13, 1120–1124, 2013 1121



RESEARCH

ARTIC

LE

Impact of Polymer Matrix on the EMI Shielding Performance for SWCNTs-Based Composites Liang et al.

Fig. 2. Representative SEM images of the cross section of SWCNT/polymer composites. (A) SCPU with 20 wt% SWCNTs, (B) ABS with 20 wt%SWCNTs, (C) epoxy with 10 wt% SWCNTs. All the samples freeze-fractured in liquid nitrogen and gold coated.

low SWCNT loadings. Figure 1 shows that the conductiv-ity of SWCNT/epoxy composites exhibited a remarkableincrease of over 10 orders of magnitude below 0.6 wt%.At 15 wt% loading, the conductivity reached 0.2 S/cm.Compared with SWCNT/epoxy composites, the SCPU andABS composites showed slower conductivity increases,resulting in higher values of the percolation threshold for

(a)

(c) (d)

(e) (f)

(b)

Fig. 3. Complex permittivity spectra of the SWCNT/SCPU (a), (b), SWCNT/ABS (c), (d) and SWCNT/epoxy (e), (f) composites with variousSWCNT loading.

SWCNT/ABS (0.599 wt%) and SWCNT/SCPU compos-ites (3.34 wt%). The percolation threshold in SWCNTcomposites is highly coupled to the properties of polymermatrix and the dispersion state of SWCNTs in the polymermatrix, as discussed below.The dispersion state of SWCNTs in the polymer matri-

ces were studied by SEM. Figure 2 displays representative

1122 J. Nanosci. Nanotechnol. 13, 1120–1124, 2013



RESEARCH

ARTIC

LE

Liang et al. Impact of Polymer Matrix on the EMI Shielding Performance for SWCNTs-Based Composites

cross-sectional SEM images of SWCNT/SCPU (20 wt%),SWCNT/ABS (20 wt%) and SWCNT/epoxy (10 wt%)composites. Compared to SWCNT/SCPU composites(Fig. 2(A)), SWCNTs were better dispersed and embeddedthroughout the polymer matrix in ABS and epoxy matrices(Figs. 2(B) and (C), respectively). Conductivity is a con-sequence of conducting path formation in the insulatingpolymer matrix. The conducting path from interconnectedSWCNT network also contributes to the EMI SE. Becauseof their better conducting networks, SWCNT/epoxy com-posites have better conductivity than both SWCNT/ABSand SWCNT/SCPU composites with similar SWCNTloadings. Unlike the thermoplastic ABS and SCPU com-posites, SWCNT/epoxy composites were prepared by dis-persing SWCNTs in the epoxy precursor 618-epoxy beforecuring with the amine-type hardener. Thus, SWCNTscould be dispersed much better due to the low viscosityof the 618-epoxy solution, leading to the formation of abetter conducting network. As a result, the best conduc-tivity, EMI performance (discussed below) was obtainedwith SWCNT/epoxy composites. Moreover, SCPU havea lower solubility than ABS due to SCPU partly cross-linking and it was harder to obtain a well-dispersedSWCNT composite. Therefore, with the same SWCNTloadings, SWCNT/ABS composites had better conductiv-ity than SWCNT/SCPU composites due to the better dis-persion of SWCNTs in ABS matrix. Therefore, SWCNT/SCPU composites had the highest percolation thresholdamong the SWCNT/polymer composites evaluated here.To evaluate the EMI shielding performance of SWCNT/

polymer composites, we measured the complex per-mittivity of the composites in the frequency range of8.2–12.4 GHz (X band). Figure 3 shows the complex per-mittivity spectra of the SWCNT/SCPU, SWCNT/ABS andSWCNT/epoxy composites containing various SWCNTloadings. Both the real (′) and imaginary (′′) permittivityincreased dramatically with increasing SWCNT loading.Furthermore, at low loadings, both the real and imaginaryparts of permittivity were almost independent of frequencyin X band. At higher loadings, the values fluctuated, per-haps because of the high electrical conductivity in thecomposites with high SWCNT loading.2 For a given load-ing, the SWCNT/epoxy composites had the highest realand imaginary permittivity and the SWCNT/SCPU com-posites had the lowest. The permittivity data pattern cor-responded well with the composite conductivity shown inFigure 1.The much-enhanced imaginary and real parts of per-

mittivity in these SWCNT composites indicated that theyare suitable for use as EMI materials in the measured fre-quency region. Figure 4(a) shows the EMI SE SWCNT/ABS composites with various loadings of SWCNTs.It can be seen that the EMI SE was almost independentof frequency. Furthermore, the EMI SE increased withincreasing of the concentration of SWCNTs. The EMISE performance of SWCNT/SCPU and SWCNT/epoxy

(a)

(b)

Fig. 4. (a) EMI SE in the range of 8.2–12.4 GHz for SWCNT/ABS (B)and composites with various SWCNT loadings. (b) Comparison of theEMI SE at frequency of 12.4 GHz among the SWCNT/SUPU (Blacksquare), SWCNT/ABS (Red circle), SWCNT/Epoxy (Blue triangle) withSCWNT loadings from 1 to 15 wt%.

composites also had similar results. As can be found inFigure 4(b), the EMI SE of SWCNT/epoxy compositeswas higher than that of SWCNT/ABS and SWCNT/SCPUcomposites. For example, with 15 wt% loading, the EMISE value for the SWCNT/epoxy composite reached 28 dBat 12.4 GHz, while the SWCNT/ABS and SWCNT/SCPUcomposites exhibited 12 dB and 9 dB, respectively. Thistrend agreed well with the conductivity (Fig. 1) and per-mittivity data (Fig. 3). The results and discussion aboveconfirm that the composites with better dispersion ofSWCNTs exhibited higher electrical conductivity, permit-tivity and better EMI shielding.

4. CONCLUSIONS

The EMI SE in the X band properties had been studiedfor SWCNT/polymer composites with three representativetypes of polymer matrices: SCPU, ABS and epoxy. Theresults demonstrated that the polymer matrix had greatimpact on the conductivity, percolation threshold, EMI SEperformance. The EMI SE of the composites was highlycorrelated with the dispersion state of the SWCNTs. Withthe best dispersion of SWCNTs in the epoxy matrix,epoxy/SWCNT composites, 28 dB EMI SE was obtainedfor the composite with 15 wt% loading of SWCNTs at

J. Nanosci. Nanotechnol. 13, 1120–1124, 2013 1123



RESEARCH

ARTIC

LE

Impact of Polymer Matrix on the EMI Shielding Performance for SWCNTs-Based Composites Liang et al.

12.4 GHz. The EMI shielding performance was stronglyrelated to the degree of SWCNT loading and with increas-ing SWCNT loading, the EMI SE of SWCNT/polymercomposites increased.

Acknowledgments: The authors gratefully acknowl-edge financial support from MOST (Grants2012CB933401, 2011CB932602 and 2011DFB50300),NSFC (Grants 50933003, 50902073 and 50903044) and100061200111.

References and Notes

1. X. F. Zhang, X. L. Dong, H. Huang, Y. Y. Liu, W. N. Wang, andX. G. Zhu, Appl. Phys. Lett. 89, 053115 (2006).

2. Y. L. Yang, M. C. Gupta, K. L. Dudley, and R. W. Lawrence, Adv.Mater. 17, 1999 (2005).

3. Y. J. Chen, M. S. Cao, T. H. Wang, and Q. Wan, Appl. Phys. Lett.84, 3367 (2004).

4. A. Wadhawan, D. Garrett, and J. M. Perez, Appl. Phys. Lett. 83, 2683(2003).

5. M. S. Dresselhaus, Nature 432, 959 (2004).6. R. Martel, T. Schmidt, H. R. Shea, T. Hertel, and P. Avouris, Appl.

Phys. Lett. 73, 2447 (1998).7. M. Zhang, S. L. Fang, A. A. Zakhidov, S. B. Lee, A. E. Aliev, and

C. D. Williams, Science 309, 1215 (2005).8. R. H. Baughman, A. A. Zakhidov, and W. A. de Heer, Science

297, 787 (2002).9. E. Minoux, O. Groening, K. B. K. Teo, S. H. Dalal, L. Gangloff,

and J. P. Schnell, Nano Lett. 5, 2135 (2005).10. J. K. Holt, H. G. Park, Y. M. Wang, M. Stadermann, A. B.

Artyukhin, and C. P. Grigoropoulos, Science 312, 1034 (2006).11. P. M. Ajayan, Chem. Rev. 99, 1787 (1999).

12. D. D. L. Chung, Carbon 39, 279 (2001).13. J. Joo and C. Y. Lee, J. Appl. Phys. 88, 513 (2000).14. E. T. Thostenson, Z. F. Ren, and T. W. Chou, Compos. Sci. Technol.

61, 1899 (2001).15. S. H. Kim, S. H. Jang, S. W. Byun, J. Y. Lee, J. S. Joo, and S. H.

Jeong, J. Appl. Polym. Sci. 87, 1969 (2001).16. H. M. Kim, K. Kim, C. Y. Lee, J. Joo, S. J. Cho, and H. S. Yoon,

Appl. Phys. Lett. 84, 589 (2004).17. Y. L. Yang and M. C. Gupta, Nano Lett. 5, 2131 (2005).18. A. Allaoui, S. Bai, H. M. Cheng, and J. B. Bai, Compos. Sci.

Technol. 62, 1993 (2002).19. R. C. Che, L. M. Peng, X. F. Duan, Q. Che, and X. L. Liang, Adv.

Mater. 16, 401 (2004).20. Z. J. Fan, G. H. Luo, Z. F. Zhang, L. Zhou, and F. Wei, Mater. Sci.

Eng. B 132, 85 (2004).21. C. S. Xiang, Y. B. Pan, X. J. Liu, X. W. Sun, X. M. Shi, and J. K.

Guo, Appl. Phys. Lett. 87, 123103 (2005).22. J. A. Roberts, T. Imholt, Z. Ye, C. A. Dyke, D. W. Price, and J. M.

Tour, J. Appl. Phys. 95, 4352 (2004).23. Y. Huang, N. Li, Y. F. Ma, F. Du, F. F. Li, X. B. He, H. J. Gao, and

Y. S. Chen, Carbon 45, 1614 (2005).24. N. Li, Y. Huang, F. Du, X. B. He, X. Lin, H. J. Gao, and Y. S. Chen,

Nano Lett. 6, 1141 (2006).25. F. Du, Y. F. Ma, X. Lv, Y. Huang, F. F. Li, and Y. S. Chen, Carbon

44, 1327 (2006).26. F. X. Li, Z. F. Liu, X. P. Liu, X. Y. Yang XY, S. N. Chen, and Y. L.

An, Macromolecules 38, 69 (2005).27. E. J. Garboczi, K. A. Snyder, J. F. Douglas, and M. F. Thorpe, Phys.

Rev. E 52, 819 (1995).28. M. B. Bryning, M. F. Islam, J. M. Kikkawa, and A. G. Yodh, Adv.

Mater. 17, 1186 (2005).29. B. Kim, J. Lee, and I. S. Yu, J. Appl. Phys. 94, 6724 (2005).30. J. C. Grunlan, A. R. Mehrabi, M. V. Bannon, and J. L. Bahr, Adv.

Mater. 16, 150 (2005).31. S. P. Obukhov, Phys. Rev. Lett. 74, 4472 (2005).

Received: 12 September 2011. Accepted: 30 November 2011.

1124 J. Nanosci. Nanotechnol. 13, 1120–1124, 2013

Documents

Impact of Polymer Matrix on the Electromagnetic ...for high performance EMI shielding.2 12 13 15 22 While most reports point out that the structure and prop-erties of CNTs strongly